This is the old United Nations University website. Visit the new site at http://unu.edu

Contents - Previous - Next

Local climatological differences between highlands and lowlands in Thailand

Masatoshi M. Yoshino

Abstract

Local differences of climate are discussed as a contribution to the establishment of highland-lowland interactive systems research and prospectives for agro-forestry development in Thailand.

First. mean climates expressed by Koppen's and Thornthwaite's classification methods are discussed. The tropical wet and dry climate (Aw) predominates in Thailand, but the tropical wet climate (Am) appears only in the southernmost parts. The semiarid climate (BS) occurs at some places in the hills and lowlands of Thailand between one and three out of 25 years. By Thornthwaite's method the semi-arid and dry subhumid conditions are also detected. Water deficit is striking in the northernmost part of the highlands in winter.

Second, distributions of some climatic elements are studied from the standpoint of local climatology. Important results obtained are as follows: The air temperature lapse rate is much greater than the normally accepted average value of 0 6 C/100 m. Frost can occur quite frequently in the highlands of the northwest. Very high temperature maxima occur frequently in April not only in the lowlands but also in the highlands. The heaviest rainfall occurs in August or September. The 24-hour maximum rainfall is 150 200 mm for the highlands but less than 100 mm for the lowlands. Soil erosion must be taken into consideration for planning agroforestry systems. Thunderstorm distribution shows a close relation to the convergence zone of upper currents, which runs meridionally through the central part of Thailand. The number of fog days and precipitation days is quite different between the NE and SW monsoon seasons and their distributions show sharp contrasts between highlands and lowlands during the prevailing seasons respectively.

Introduction

In order to discuss highland-lowland interactive systems from the viewpoint of climatic conditions, the present paper describes various aspects of differences in local climates.

One of the comprehensive climatographies for Thailand is Climate of Thailand. prepared by the Department of the Air Force, USA (1965). Although this publication was written for the Air Force services, much climatological information for highlandlowland agro-forestry can be derived from it. A brief but very good description of problems in the study of climatology of Thailand was given by Dhararaks (1959). One chapter was dedicated to climatography of Thailand as a part of a regional geography by Pendleton (1962) and contributes a good general view of climate and weather.

Kyuma (1971) applied the climatic classification method of Thornthwaite (1948) to Thailand as part of the South and Southeast Asian climate studies and pointed out that inner Thailand is the driest area in Southeast Asia and, on this account, grumusols occur there. Kyuma (1972) further studied the climatic classification of South and Southeast Asia, including Thailand, and introduced a new method. Distributions of rainfall, soil moisture storage, water deficiency, and water surplus calculated by Thornthwaite's 1948 method for Thailand were drawn for every month by Maruyama (1978). According to his studies, water dificiency in February, March. and April is striking in the continental part of Thailand, except for the highlands. On the other hand, soil moisture storage reaches the high values of more than 100 mm in the mountain areas in the northern part of Thailand in July. August, and September.

In the present paper. the ''climatic year'' or ''year climate'' expressed by Koppen's and Thornthwaite's classification methods is discussed first, and distributions of some climatic elements, which are important for land-use planning in the context of highland-lowland and agro-forestry interactive systems research, are dealt with later on.

Local Differences of Year Climate Expressed by Koppen's Method

As is well known, Koppen's climatic classification is applicable on the basis of the long-term yearly mean values of air temperature and precipitation. Russell (1932; 1934), however, applied the Koppen method to illustrate the climatic conditions for a single year using the single-year values. He defined this as ''climatic year'' and studied the ''climatic year'' boundaries between dry and humid climates, or between C and D climates in the United States. It became clear that the "climatic year'' boundaries shift position from year to year. Sekiguti (1951) also applied the Koppen classification method to the single-year values during 1 896 1940 in Japan, defining it as "year climate.'' He demonstrated rather different distribution patterns of "year climate'' for each year and variations in the frequencies of various climatic types at each station. Mizukoshi (1971 ) studied the regional divisions of Monsoon Asia according to the same method based on data from 450 stations.

In the present paper, adapting the Koppen classification method to assess the climatic conditions for one year only. the climatic type for that year is determined and called the "year climate." The year climate was determined at 46 stations in Thailand for the period 1951 to 1975.

TABLE 1. Frequency of Year Climate by Koppen's Climatic Classification at Each Station for the Years 1951 - 1975

The results obtained are shown in Tables 1 and 2. The distributions of occurrence frequencies of Aw, Am. Af, BS, and Cw climates are given in Fig. 1. Here, the characters of the climatic types of Koppen are as follows:

A = air temperature of coldest month above 18 C,
B = evaporation exceeds precipitation,
C = coldest month between 18° C and 0° C,
f = no dry season; driest month precipitation over 60 mm.
m = no dry season; driest month precipitation less than 60 mm,
w = winter dry; at least 10 times as much rain in wettest month of summer as in driest month of winter,
S = steppe climate; r = (1 ~ 2)t in winter wet climate,r = (1 ~ 2) (t+14) in summer wet climate, and r= (1 ~ 2) (t+ 7) in no dry season climate, where r = annual precipitation in cm and t = annual mean temperature in degrees Celsius

The tropical wet and dry climate (Aw) predominates throughout the entire area except for the southern part, as shown in Fig. 1. In the highlands of the western, northwestern, and northern regions of Thailand, the occurrence frequency is rather high: 22 to 24 out of 25 years. The southern part of Thailand has the tropical wet climate with the driest month precipitation less than 60 mm (Am climate).

According to the Koppen classification based on the longterm yearly mean values, the dominant distribution areas of Aw and Af climates are almost identical, but there is a spot of Af climate on the peninsular coast of the Gulf of Thailand (Mizukoshi 1971) .

TABLE 2. Frequency (%) of Year Climate by Koppen's Climatic Classification for Each Year, 19511975

It is noteworthy that the semi-arid (steppe) climate (BS) appears in the lowlands, even though its frequency is only 1 to 3 out of 25 years. It does not appear on the Koppen classification based on long-term yearly mean values. These conditions cannot be neglected in agro-forestry planning. In Chiang Rai the humid mesothermal winter dry climate (Cw) appears 4 out of 25 years. Cw climate covers large areas of China and extends occasionally into the northernmost part of the Thai highlands during cold winters. This fact might also be important to agro-forestry in these regions. As Trewartha (1954; 1961) pointed out. Cw climates appear in two characteristic locations: (a) in tropical highlands where, because of altitude, the temperature drops below that of the surrounding lowlands with Aw climate; and (b) in mild subtropical monsoon lands. For northern Thailand the former case is assumed.

Local Differences of Year Climate Expressed by Thornthwaite's Method

An adaptation of the climatic classification of Thornthwaite (1948) can be made by using long-term yearly mean values of both monthly mean air temperature and monthly precipitation. However, water balance can also be calculated for a single year using monthly evapotranspiration and precipitation values for a particular year. The moisture index. ‰ is expressed by: Im = Ih - 0.6Ia where Ih and Ia are indices of humidity and aridity respectively. Ih= 100s/n and Ia = 100d/n, where s is water surplus, d is water deficiency, and n is water need. Thornthwaite showed the various climatic types together with their limits as given in the upper part of Table 3. Since the computation of water balance for every year from 1951 to 1975 at many stations takes too much time, we computed the water balance only at the 12 stations in the northwestern part of Thailand and one arid station in the south for contrast. The distributions of occurrence frequency of more humid climate (B₄ + B₃ + B₂), less humid climate (B₁). moist subhumid climate (C₂), dry subhumid climate (C₁). semiarid climate (D). and arid climate (E) are shown in Fig. 2.

Against expectation, dry subhumid climate appears between 10 and 14 out of 25 years at most stations; only Chiang Rai has a prevailing humid climate and Mae Hong Son (271.7 m above sea level) on the northwestern border, a moist subhumid climate. It is a rather startling fact that semi-arid (D) climate prevails with the occurrence frequency of about 40 per cent at Lampang (241.6 m), Phrae (158 m), and Tak (115.6 m), which are located in the central part of the northwestern region of Thailand. Using only the Koppen approach. D climate is given only for Tak and C' climate for Lampang and Phrae, as shown in Table 3. Such a dry condition has been detectable in Table 1 and in Fig. 1 (BS climate distribution), but is clearly seen by this year climate through adaptation of the Thornthwaite method.

FIG. 1. Distributions of Occurrence Frequency of Aw, Am, Af, BS, and Cw Climates According to the Year Climate by Koppen's Method. Unit: number of years during 1951-1975

The climatic subdivisions are defined in terms of the aridity index, la and humidity index, Ib as follows: Moist climate (A. B, C2) Aridity index. Ia r little or no water deficiency 0 - 16.7 w moderate winter water deficiency 16.7 - 33.3 w2 large winter water deficiency 33.3 +

TABLE 3. Frequency of Year Climate by Thornthwaite's Climatic Classification (/m) at Each Station for the Years 1951 - 1 975

*Obtained by the long year mean values ( Kyuma 1971 )

TABLE 4 Frequency of Year Climate by Thornthwaite's Climatic Classification (/a or Ih) at Each Station for the Years 1951 - 1975

The most striking features shown in Table 4 are: (a) moderate winter water deficiency (w of Ia) occurs quite frequently at Chiang Rai and Mae Hong Son and large winter water deficiencies at Mae Sot; (b) little or no water surplus (d of Ih ) appears frequently at Lampang, Phrae, Tak, and Phetchabun. but moderate summer water surplus predominates at Mae Sariang, Uttaradit, Phitsanulok and other stations: (c) large water surplus in summer (w2 of Ih occurs 1 or 2 years out of 25 years which could be detected only by the year climate method; (d) for example. at Chiang Mai the occurrence of w of the aridity index, /a and the occurrence of w of the humidity index, Ih. are both 7 years. and w2 of Ih is 5 out of 25 years. This means that various conditions occur year to year. In conclusion, the summer water surplus appears not so often in northwestern Thailand. Except for certain small areas, the surplus is moderate and the frequency is less than 30 percent. On the other hand, moderate winter water deficiency occurs with a frequency of 25 to 40 per cent, or more, in the highlands of northernmost Thailand, as shown in Fig. 3.

Local Differences of Some Climatic Elements

Air temperature lapse rate

Local differences in air temperature occur markedly in accordance with the altitudinal change of the topography. The temperature decrease with height (lapse rate) has an average value of 0.6- C/100 m, but its deviation is great. It changes from region to region, season to season, and elevation to elevation Therefore, it is risky to estimate the lapse rate for unknown regions (Yoshino 1975).

The lapse rates were calculated by using the 14 stations in northwestern Thailand as shown in Table 5. It is generally greater than the average status of 0.6° C/100 m. The annual mean temperature lapse rate is 1.5° C/100 m for the region studied with an elevational range from 60 m to 450 m above sea level The lapse rate of mean maximum temperature in July shows a very high value of 2.9 C/100 m. It is difficult to say to what elevation this lapse rate can be extrapolated. but it is nevertheless important to estimate temperatures in the mountain regions for agricultural and horticultural purposes. Although we have no knowledge of the extrapolation reliability to the higher regions, the air temperature must be lower than that anticipated from the general lapse rate for regions higher than 500 m above sea level.

FIG. 2. Distribution of Occurrence Frequency of B, C, and D Climatic Types According to the Year Climate by Thornthwaite's Method. Unit: number of years during 1951-1975

TABLE 5. Air Temperature Lapse Rate ( C/100 m) in Northwestern Thailand

Frost occurrence

At Loei (252.7 m above sea level) the observed record of the lowest extreme minimum temperature during the seven years was 0° C. Among the stations with altitudes above 100 m above sea level in northwestern Thailand, ten have records lower than 4.4° C during the available periods of observation which range from 7 to 24 years. A screen temperature of 4° C normally means the occurrence of ground frost at the observation point and, if the local irregularities of air temperature caused by microtopographical conditions are taken into consideration, there must be frost dangers at many spots in the surrounding areas. It has been reported that 2° C was recorded at a station at Huai Thung Choa (98°40'E, 17°10'N, 1.200 m above sea level) during the last two years (PisitFrost occurrence in the mountain regions of the tropics has been dealt with for Sri Lanka (Domros 1970; 1974), South India (Lengerke 1978), and Java (Domros 1976). In Sri Lanka frost is expected to occur in about every fifth year in the depressions and hollows at altitudes above 1.400 m and an annual average of 1.5 days of frost in the air and 8.7 days of ground frost were reported at an altitude of about 1,890 m. On the plateau in South India, frost occurs commonly above 1.800 m. As has been mentioned above. the minimum temperatures in northwestern Thailand must be lower than the values generally expected. Therefore, frost occurrence should by no means be a rare phenomenon. Its frequency must be studied statistically in detail in the future.

FIG. 3. Distribution of Occurrence Frequency of w and w2 Climate of Aridity Index, and d, w, and w2 Climate of Humidity Index, According to the Year Climate by Thornthwaite's Method. Unit: number of years during 1951-1975 Voraurai, personal communication).

Extreme maximum temperatures

Extremely high temperatures occur in April. The almost vertical position of the sun and the battle by the NE monsoon to retain itself just before the beginning of SW monsoon make it the hottest month of the year in Thailand. The maximum temperature recorded for Thailand is 45.6 C at Bangkok. In northwestern Thailand, the record is 44.4° C at Uttaradit (63.3 m above sea level) and 43.9° C at many stations at every altitude up to about 300 m above sea level. Thus, very high maximum temperature in April. say higher than 38° C, can be expected frequently and must be taken into consideration for any horticultural planning.

Rainfall intensity

In the tropics. heavy rainfall during a short period is an important factor for land use. field cultivation, construction, and so on. In northwestern Thailand, unfortunately, there are few data on rainfall intensity for short time duration.

The distribution of observed records of the 24-hour maximum rainfall is shown in Fig. 4. The 24-hour maximum rainfall is more than 200 mm in the western mountains around Maelamun and in the central part of the northern borderlands (south of Doi Luang). In the northern part of northwestern Thailand, around Chiang Rai, it is between 150 and 200 mm. The mountain stations at Huai Thung Choa mentioned above recorded 175.5 mm (15 September 1977) and 121.7 mm (1 November 1977) of daily rainfall during the last one and a half years. The lowlands in the valleys between them have a value lower than 100 mm. The values of 48 mm at Lampang (241.6 m above sea level) or 63.5 mm at Tak (115 6 m) show a sharp contrast to that of 210.8 mm at Mae Sot (210.8 m), situated in the western mountains near the Burmese border. These maxima occur in August or September in the northwest in association with strong convective currents under the influence of the SW monsoon. Therefore, the stations at the western foot of the mountain ranges such as Mae Sot record the greater vaIues

FIG.4. Distribution of Observed Records of 24-hour Maximum Rainfall

FIG. 5. Relation between the Observed Records of 1hour and 24-hour Maximum Rainfalls for Every Month Values at Chiang Rai and Chiang Mai

The 24 hour maximum rainfall is greater in the southern part of Thailand. The observed records are 302.3 mm at Sattahip and 337.8 mm at Chanthaburi on the coast of the Gulf of Thailand in October and 342 9 mm at Ban Don and 375.9 mm at Songkhla in the Malay Peninsula in November or December.

The data for 1-hour maximum rainfall are available only at a few stations. Fig. 5 shows the relation between 24 hour maximum and 1-hour maximum rainfall based on the data for every month at Chiang Mai and Chiang Rail The ratio of the 1hour maximum to the 24-hour maximum is about 1:4 on the range 0 to 50 mm of 24-hour maximum. 1:3 on the range of 50 to 100 mm. and 1:2 on the range of 100 to 200 mm This means that the annual march of the rainfall intensity is quite obvious: heavy 1-hour rainfall is expected during the late SW monsoon season and very slight during the NE monsoon season for northwestern Thailand. Serious erosion must occur in the highlands of this part of Thailand because of high intensity rainfall, because the ground surface is almost saturated by heavy rainfall after the onset of the SW monsoon season and the highest intensities come at the end of the SW monsoon season.

Thunderstorms

In April and May thunderstorms are quite numerous. They occur usually in the afternoons and early evenings. The axis of maximum thunderstorm occurrences is located in a north-south direction from Loei to Korat to Chanthaburi on the Gulf of Thailand in March. April. and May. In May, the numbers of days with thunderstorms are at a maximum for almost all stations: that is, 26 to 27 along the axis as shown in Fig. 6.

It is difficult to explain the reasons why the axis runs meridionally in the central part of Thailand. One reason suggested is that the convergence zone between northeast currents and southwest currents play an important role. As the convergence zone runs from northern Vietnam to the Gulf of Thailand crossing Laos (Yoshino 1969). the convective type thunderstorm occurs frequently along this convergence zone.

On the other hand, Dhararaks (1959) wrote that the convective thunderstorms occur in the afternoon because of the development of a thermal low over central Thailand in March. April. and May Southeast or south winds, strengthened by the trade wind over the South China Sea. blow from the Gulf of Thailand into this thermal low, combining with an afternoon sea breeze in the coastal region This type of wind with fine weather is locally called the Lom wao (wind for kite).

Rain showers often occur along the local front formed between the warm afternoon air at the lower level and the cold NE monsoon currents in the coastal region from November to May, especially in January and February. This shower activity is important because it produces moisture during the dry season. Local people call it ''Mango shower'' Further studies of the mechanism of the ''Mango shower'' together with the Lom wao from the standpoint of local, synoptic meteorology are needed

Fog days and precipitation days

Fog is quite frequent during the NE monsoon season and, in contrast. very rare during the SW monsoon season Fig 8 shows the distribution of the number of fog days In the western mountain region, it reaches 100 days during a period of five months with a maximum of 123 days at Mae Sot. At Lampang, however. it is only 35 days This contrast in northwestern Thailand is quite striking. In the highland region surrounding the north, the average is about 80 days.

FIG. 6. Distribution of Thunderstorms in May

FIG. 8. Distributions of Number of Fog Days

During the SW monsoon, the number of fog days is less than 10 also during a period of five months, except for the northern border highland of central Thailand. In Bangkok, fog occurrences are more frequent than in the adjacent areas during both the SW and NE monsoon seasons, probably because of the city effect. In the peninsula. Chumphon on the coast of Gulf of Thailand has higher values than Ranong or Phuket to the west.

The number of precipitation days shows the opposite tendency to that of fog days. Namely, a high frequency during the SW monsoon season and lower during the NE monsoon season as shown in Fig 9. Therefore, during the SW monsoon season, the number of precipitation days is high along the western foot of mountain ranges and low along the eastern foot. For example. a sharp contrast is seen between Mae Sot (113 days) and Tak (45 days). In other areas. almost all northwestern Thailand has about 60 to 80 precipitation days during the five-month SW monsoon season. This means that rain falls on about half the total number of days which is smaller than that generally expected for Monsoon Asia.

Nakajima (1975) calculated the standard deviations of monthly precipitation and the precipitation variability for every month. These values show a marked contrast between the SW monsoon season and the NE monsoon season similar to that shown by the number of precipitation days.

Conclusion

Summarizing the results discussed above. it can be said that the following climatic conditions must be taken into account in the development of any highland-lowland interactive systems research and agro-forestry initiative in Northern Thailand. These include: water deficiency; distributions of air temperature and rainfall; frost occurrence; extreme high temperature; rainfall intensity; as well as fog and dew.

The future problems to be studied are: (i) the observation of air temperature, rainfall. wind, etc. from the standpoint of topoclimatology, (ii) rainfall intensity observation with a dense network in the mountain region itself, (iii) watershed management survey such as rill and gully formation due to heavy rainfall from a micro-geomorphological viewpoint and (iv) measurement of debris amount and suspension amount in the rivers in association with run-off observations.

FIG. 9. Distributions of Number of Precipitation Days

The analysis of the data accumulated at the proposed stations in the mountain regions is also needed in order to clarify the highland conditions.

References

Department of the Air Force 1965. Climate of Thailand 1st Weather Wing. Detachment 51, APO San Francisco 96525.

Dhararaks, C.K.V. 1959. ''A Note on the Climatic Condition of Thailand'' Proceedings of 9th Pacific Sci. Congress (Meteorology). 13, pp. 79-80.

Domros, M. 1970 ''Frost in Ceylon '' Arch. Met. Geophy. Biokl (B). 19 43 52

____________1974. ''The Agroclimate of Ceylon '' Ecological Res. 2: 1 - 238.

____________1976. ''Über das Vorkommen von Frost auf Java/Indonesien. insbesondere in den Pengalengan Highlands.'' Erdkunde. 30: 97 108. Kyuma. K. 1971. ''Climate of South and Southeast Asia According to Thornthwaite's Classification Scheme.'' Southeast Asian Studies) Kyoto(.9 (1): 1 36 1 58.

____________1972. ''Numerical Classification of the Climate of South and Southeast Asia '' Southeast Asian Studies (Kyoto) 9 (4): 502 - 521

Lengerke, H J. von 1978. ''On the Short-Term Predictability of Frost and Frost Protection.'' Agricultural Met. 19: 1 - 10.

Maruyama. E 1978 ''Fluctuation of Paddy Yield and Water Resources in Southeast Asia.'' In K. Takahashi and M M Yoshino, eds, Climatic Change and Food Production. pp 155-166. Tokyo: University of Tokyo Press.

Mizukoshi, M. 1971. ''Regional Divisions of Monsoon Asia by Koppen's Classification of Climate.'' In M.M. Yoshino. ea.. Water Balance of Monsoon Asia. pp. 259 - 273. Tokyo: University of Tokyo Press

Nakajima, Ch. 1975. ''The Climate of Southeast Asia.'' 2, ''Heavy Rainfall in Laos, Thailand, Malaysia and Singapore.'' Southeast Asian Studies (Kyoto). 13 (2): 308 336. (In Japanese with English abstract.)

Pendleton. R.L. 1962. Thailand Aspects of Landscape and Life. pp. 1-321. New York: Duell. Sloan and Pearce.

Russell, R.J. 1932. ''Dry Climates of the United States,'' Part II. "Frequency of Dry and Desert Years 1901-1920.'' Univ. of Calif. Publications in Geog.. 5 (5): 245 274.

____________1934. ''Climatic Years." Geog. Rev . 24: 92-103

Sekiguti, T. 1951. ''On the Year Climate in Japan.'' Geog. Rev. of Japan. 24 (6): 175-185 (In Japanese with English abstract.)

Thornthwaite, C.W. 1948. ''An Approach toward a Rational Classification of Climate.'' Geog. Rev. 38: 55-94.

Trewartha, G T. 1954. An Introduction to Climate New York: McGraw-Hill.

____________.1961. The Earth's Problem Climates. pp. 1 - 334. Madison: University of Wisconsin Press.

Yoshino, M.M. 1969. ''Intertropical Convergence Zone and Polar Frontal Zone over South. Southeast and East Asia: a Climatological View.'' Annalen der Met.. N F. 4: 212-220. Also in Climatological Notes. 1: 1 - 71.

____________1975. Climate in a Small Area. Tokyo: University of Tokyo Press.

Acknowledgements

I would like to thank Mr. Visit Rasmidatta for sending me the air temperature and rainfall data for 1951 - 1975. Mr. H. Kurosaka for computation and determination of climatic types by Koppen and Thornthwaite methods, Mrs K Masuda for drawing the figures, and Miss J Tanabe for typing the manuscript.

Contents - Previous - Next